* feat(native): admit raw binaries with caller-supplied architecture and entry
* fix(native): honor raw-binary preparation and entry identity
* fix(ghidra): stream raw snapshot attestation with cancellation
* fix(native): preserve and validate raw lifecycle entries
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* test(ida): expect a missing procedure to be reported as invalid input
* test(ida): name the invalid_request code and its category for a missing procedure
* fix(ida): report a procedure IDA cannot resolve as invalid input
* test(ida): pin the missing-procedure issue shape and the lookups that stay unreadable
* fix(ida): treat a lookup answer for another query as a producer fault
* test(ida): expect an argument-independent issue path and pin legacy error replies and escaping
* fix(ida): report a missing procedure without naming an argument path
* test(ida): expect provider-derived selectors to stay producer faults in composed workflows
* fix(ida): keep provider-derived procedure selectors out of the caller's input errors
* test(ida): cover session forwarding, parameter hygiene and structured derived-node errors
* docs(ida): describe when procedure resolution blames the caller
* feat(harmony): project HarmonyOS package inventory evidence
First HarmonyOS increment from #1431: HAP, HSP, and App Pack roots are
classified through the hardened ZIP reader and projected into an
execution-free application graph with packaging-model hints, exact
component paths and hashes, App Pack children by path, runtime-family
hints, and path-based N-API bridge hypotheses. HarmonyOS .har stays
unclassified by suffix because .har also names HTTP Archive JSON.
Adds the project_harmony_application_graph tool (CLI and MCP), synthetic
composition coverage, a pinned real-package verify lane, and docs.
* refactor(domain): share application graph projection scaffolding
The HarmonyOS projection mirrors the Android and Apple projections by
design, which pushed jscpd duplication over its 2.5 percent budget.
Extract the identical pieces (inventory input schema, projected
component schema and occurrence flattening, bridge candidate schema
factory, coverage assembly, comparator, canonical digest) into
domain/mobileApplicationGraphProjection.ts and route all three
projections plus the composition fixtures through it. Duplication drops
from 2.52 to 2.49 percent with no behavior change: projection ids,
classifications, and limitations are byte-identical.
* fix(harmony): address review, stable app-pack roots, and rotated fixture
- rebased onto main; the shared projection scaffolding now keeps main's
nativeRuntimeConvention loop and the harmony runtime families adopt the
same react-native/flutter/unity path conventions
- the .app App Pack suffix is now a root-input rule only (verified ZIP
magic in classifyRootFormat and BinaryTargetResolver); nested .app ZIP
members inside other packages keep the generic zip classification, with
a regression test, so existing inventories do not change
- the signing limitation now states that an empty signing list does not
mean unsigned because packaging-tool signatures live in the ZIP signing
block, and the app-pack root rule is disclosed as suffix-based
- repinned the real-package fixture to the current VHome 2.6.14-beta asset
after the upstream deleted v5.0.111-beta and replaced the v2.6.14-beta
asset within hours; the pin records both observed digests
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* fix(bun): report the real runtime and enforce worker memory budgets on Windows
Under Bun, REA recorded Node/V8 identity it did not observe and its JavaScript helper processes ran without a memory bound, because Bun ignores the V8 heap flags REA passes. Source-map trace Evidence claimed runtime node-v8 with Bun's synthetic V8 heap limit (the decoder then rejected that limit, so trace_web_source_location always failed), analysis resource failures reported heap limits Bun does not enforce, the HAR decoder claimed a fixed 192 MiB heap, and doctor reported Bun's Node.js compatibility version as the Node version.
Identify the runtime once from process.versions. Source-map reports record id bun with the Bun and Node.js compatibility versions and a null v8_heap_limit_bytes; the decoder checks the heap budget only for V8 reports and rejects a report from a runtime other than the launching one. The analysis worker reports a null heap limit under Bun, HAR limitations and failures omit heap sizes under Bun, and the doctor node check reports Bun and passes when its Node.js compatibility version is supported.
On Windows under Bun, each helper's owned Job Object now carries a commit limit of its heap budget plus a measured 512 MiB Bun runtime reserve (704 MiB for the codec and HAR decoder, 1536 MiB for the default analyzer). The native addon accepts an optional page-aligned job memory limit on process_spawn, verifies it by readback, associates a private completion port before the child is created, and reports a sticky memoryLimitReached and peak job commit on each poll; inspect reports jobMemoryLimit. A requested limit with an artifact lacking the control refuses the launch, and non-Windows launchers refuse a limit.
Reaching the limit is classified by the job notification rather than Bun's exit status: the analyzer fails the current source with a memory resource failure reporting worker_process_memory_limit_bytes, the codec returns a memory resource failure, and the HAR decoder returns its memory failure with process_memory_limit_bytes. Bun source-map Evidence records process_memory_limit_bytes, verified against the launcher. Node.js launches and outputs are unchanged.
* fix(windows): drain memory notifications after observing exit
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Setup run through bunx or bun x registered the bin inside bun's temporary
package cache. Detect bun's package-runner launch from npm_command=exec
with a bun user agent and register the pinned bunx launcher, wrapped in
cmd.exe on Windows like the npx launcher. Doctor, update maintenance and
uninstall recognize it as REA-owned, and setup's printed commands name
the runner that launched it.
Read the runner variables case-insensitively on Windows, where the CLI
environment snapshot upper-cases variable names and setup run through
npx also fell back to the cache path.
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(javascript): recover code-split bundles from multiple inputs
recover_javascript_sources accepted a single path, and
validateModuleSet bound every module's provenance to that one snapshot.
A code-split app (runtime + chunks, or an Electron main/preload/renderer
set) had to be recovered one file per call, and modules recovered alone
could not resolve imports across chunks: cross-chunk require(id)
references stayed as numeric ids with property reads.
Add an optional additional_paths input. The listed files are
concatenated, in order after path, into the single engine snapshot, so
one Wakaru run sees every module and cross-chunk imports resolve.
validateModuleSet's single-snapshot binding is unchanged. Module
provenance is attributed back to the input file its first extraction
range starts in, with byte ranges rebased to that file's own offsets;
a module whose ranges cross an input boundary keeps the concatenated
snapshot identity and offsets, recorded as an explicit limitation.
Single-path behavior is byte-for-byte unchanged (one segment,
offset 0).
Fixes#1605
* fix(javascript): preserve source identity in multi-input recovery
---------
Co-authored-by: JingHao-Leon <w15909130000@gmail.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(update): update Bun global installations
`rea update` only recognized npm global installations and looked up
releases with `npm view`, so a `bun add --global rea-agents` install on a
machine without npm failed with `spawn npm ENOENT`.
Detect the running package inside Bun's global directory
(BUN_INSTALL_GLOBAL_DIR, $BUN_INSTALL/install/global,
$XDG_CACHE_HOME/.bun/install/global, ~/.bun/install/global), resolve the
latest release with `bun info`, install with
`bun add --global --ignore-scripts`, then verify the installed entry point
and plan integration maintenance exactly as for npm.
Bun commands run inside the owning global directory with
BUN_INSTALL_GLOBAL_DIR pinned, so the caller's current project is never
selected. A global directory without package.json is not treated as an
owned installation: Bun then installs into the nearest ancestor project.
npm installations keep their existing commands and messages.
* fix(update): gate Bun self-updates on supported CLI versions
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Use the existing JSON-text status checker for explicit package.json imports while retaining separate package-metadata validity. Preserve malformed/unavailable status and partial facts after cancellation.
Co-authored-by: Immanuel Tikhonov <122638311+immanuwell@users.noreply.github.com>
Use the canonical current and supported legacy process-provider rule for authenticated capture Evidence at both CLI and MCP admission boundaries. Preserve successful Evidence and distinguish provider input errors from digest authentication failures.
Closes#1757.
Stream legacy snapshot cache rows independently, authenticate complete Evidence once, and restore exact immutable payloads without assembling duplicate full graphs. Preserve the existing file format, provenance, resource admission, cancellation and cache export isolation.
Verified through real Linux SDK MCP import of the original 83 MB Ghidra snapshot and full 41 MB bundle hash, actual Windows CLI/MCP PE recovery and the installed Windows large-file decoder boundary, focused tests and static checks.
Fixes#1680. Refs #1065.
Seal file-backed MCP imports at the session boundary while preserving mutable
ordinary file reads. Validate internally constructed ledger candidates directly
with the existing immutable bundle constructor.
Assisted-by: Suika (gensokyo-skills:suika-scatter-gather-planning)
Fuse upstream JSON parsing, scalar packing, heap admission, and value assembly into synchronous token stages while retaining asynchronous file reads and existing capacity, cancellation, decoding, identity, and typed-error semantics. Declare stream-chain 4.2.6 directly.
Validation: 30 focused tests; check:fast, Knip and formatting; actual Flameox paired full-content oracles; real Linux evidence replay and Windows source-built package/Ghidra snapshot recovery; final-head CI 31/31 success.
Refs #1065, #1680.
* fix(build): emit no runtime imports for type-only bindings
Under verbatimModuleSyntax, an import whose specifiers all carry inline
`type` modifiers compiles to `import {} from "..."`, a side-effect import
that still resolves and evaluates the module. 78 compiled modules carried
one, including `import {} from "playwright-core"` in the browser scenario
session, which loaded Playwright's module graph at MCP startup.
Use top-level `import type` for these bindings and enable oxlint's
typescript/no-import-type-side-effects rule so the pattern cannot return.
Refs #1685
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* perf(mcp): load Playwright when a scenario launches, not at startup
The browser scenario and Electron active providers are constructed before
the stdio server serves, and both bound Playwright's launchers at module
load. Resolve `chromium` and `_electron` from playwright-core when a
capture opens instead, as the web module resolver already does.
MCP startup now resolves 1,312 modules (4,902 resolutions) instead of
1,598 (6,076) before reading the first request, and no longer loads
playwright-core. Provider construction, the tool catalog, and capability
inventory are unchanged.
A boundary test spawns the stdio server, initializes, lists tools, and
asserts playwright-core and isomorphic-git were not resolved; the CLI
probe now exercises the real launch path to prove detection.
Closes#1685
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(evidence): discover retained records without full payloads
* fix(evidence): guide oversized results to retained discovery
* style(tests): format retained discovery recovery assertion
Select the shallowest inventoried nondependency manifest and retain dependency package evidence without assigning Electron entry roles. Keep the artifact root when no application manifest is observed.
Adapted from PR #1392, with compiled CLI and stdio MCP regressions for root selection and dependency-only inputs.
Co-authored-by: Marnix Pluim <marnixpluim@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Build ordered per-procedure indices from admitted nodes instead of scanning every node twice for each call binding. Preserve edge order, capacity limits, omitted-parameter unknowns, and surviving return bindings.
Validation: five focused service tests and check:fast passed. Four paired Flameox blocks at 4096 call sites passed all eight full-result digest oracles. Median paired reduction was 183.02 ms (71.4%) on this synthetic fan-out workload; real-provider performance is unverified.
Flameox evidence: 5a4c98ed7d224587c10a0f4596c8680d3a0330eb112a629014448ff1ea4634e7.
Replace a graph-wide node scan for each parameter use with membership in the set of admitted node IDs. Preserve node-budget omission, pagination, and existing ID-collision behavior.
Flameox profiles identify the eliminated scan. Six-run medians on a bounded synthetic four-procedure workload improved from 2538.17 ms to 220.32 ms (11.52x); full JSON responses have identical SHA-256 4071fc7ec17be01e6a23d53814a47b3b1bb0ebd80099d21ef8431da487801725. This measures production trace composition with a synthetic provider, not real Ghidra target speed.
External design research through DeepWiki was checked against Ghidra 12.1.4 DataGraph: https://github.com/NationalSecurityAgency/ghidra/blob/8b6bbb857accdfa20dc5b2f5dea471178c2e9fbc/Ghidra/Features/CodeCompare/src/main/java/ghidra/features/codecompare/graphanalysis/DataGraph.java
Require a caller-owned artifact resource scope for source traversal and Git metadata reads. Preserve cancellation, filesystem reasons, partial entries and bytes, and cleanup observations through import and the CLI. Transfer exact failed file owners at regular-file admission and close boundaries, including multiple failures encountered by Git reference resolution.
Validation: 169 focused tests passed across 26 files; two Windows-only cases skipped. Build, typecheck, lint, formatting, knip, architecture guards, discovery and docs checks passed.
BREAKING CHANGE: Reference source readers and imports require an ArtifactResourceScope. VCS metadata reads return Result, and CLI import failures include observed details and partial evidence.
Build package, JSON-module and source-map indexes once instead of rescanning every observation list per file. Preserve first-observation lookup semantics when paths repeat and share each JSON lookup with the file projection.
Validation: duplicate-observation consumer regression and reconstruction suites. Flameox captured before/after CPU profiles on 3,000 JSON files; the complete CLI output stayed byte-identical. Exploratory samples do not establish an end-to-end timing improvement.
Require the caller artifact resource scope when reading JavaScript artifacts. Transfer failed nested ASAR cleanup to that scope so later work retries the same owner before admission. Preserve traversal failures and attempt every remaining iterator during unwind.
Validation: reconstruction, integrity, content identity, cancellation and nested-reader cleanup boundary checks.
BREAKING CHANGE: readJavaScriptArtifactFiles now requires ArtifactResourceScope as its third argument; the optional abort signal is fourth.
Require an ArtifactResourceScope for web script exports and share its lifetime with CLI or MCP shutdown. Retain setup-failed output trees in firmware and Wakaru providers. Preserve completed publications when descriptor cleanup fails, returning the full result after a successful retry or as a partial observation while cleanup remains unresolved.
Keep failed reader, root descriptor, and output-tree cleanup with the caller. Drain admitted work before close, retry the same retained owners, and block new resource work while cleanup is incomplete. Preserve primary failures and complete inventory observations.
Thread ownership through JavaScript reconstruction, the CLI's one-shot scope, MCP server shutdown, verifier scripts, and test fixtures. Remove redundant ownership flags and the unused validated test wrapper.
Validate with 850 passing focused tests across 90 suites (two skipped), targeted classification and cleanup retries, static and unused-code checks, and real macOS CLI/MCP directory, ZIP, and DMG workflows. Native Windows and deep-analysis providers were not verified.
BREAKING CHANGE: Direct artifact inventory/extraction and JavaScript analysis APIs require an explicit caller-owned ArtifactResourceScope. CLI and MCP input contracts are unchanged.
Inspect each entry without following links before applying directory-only patterns. Retain the winning rule and original entry path while pruning excluded directories before reading their children.
Validate directory exclusions, negation, symlinks, explicit secret rules, and compiled CLI provenance alongside reference reader failure and cancellation coverage.
Record the exact matching pattern and distinguish project, default, caller, and configured-sensitive exclusions without changing ignore ordering. Include that provenance in graph commitments and validate the contract through compiled CLI and MCP consumers.
BREAKING CHANGE: Pattern-based reference graph exclusions require pattern and distinguish project-ignored and default-ignored reasons. Re-import saved graphs that omit matching patterns.
Pass typed workflow results directly to createEvidence, which validates generic JSON and takes the owned snapshot. Remove the extra generic parse and full result clone from all five workflow helpers; domain producers retain their specific schema validation.
Flameox confirmed JSON validation and copying costs. Alternating warm runs creating Evidence for a valid 2,001-node trace reduced median boundary time from 15.94 ms to 11.20 ms over 120 samples per variant, with identical output digests and independent mutable-input snapshots. This does not measure end-to-end workflow latency.
Validation: 39 focused CLI, SDK MCP and filesystem workflow tests and 17 Evidence tests passed; build, static and documentation checks passed. No deep analysis providers were exercised for this change.
Share root filesystem failure classification between import preflight and traversal. Missing roots remain invalid selections; permission and I/O errors retain their reason and reach the CLI as execution failures. Preserve cancellation priority and expose unexpected exceptions.
Validation: compiled CLI permission regression, direct reader error details, import/traversal and unreadable-directory boundaries passed. Windows-only cases were skipped on macOS; static, architecture and documentation checks passed.